An Empirical Approach to a Standard Practice: Drug Testing
Previous research has examined the impact of drug testing frequency on a variety of post-release outcomes (e.g., Grommon et al., 2013; Haapanen & Britton, 2002; Hawken & Kleiman, 2009; Hicks et al., 2020; Kilmer, 2008; Lattimore et al., 2016), but there has been limited exploration of how drug testing frequency impacts positive test rates (Kilmer, 2008). Some exceptions include studies examining those incarcerated in prison (Nguyen et al., 2021) and juvenile parolees (Haapanen & Britton, 2002); however, a gap remains in our understanding of how the testing rate may be related to the positivity rate among individuals on probation, the largest group of people under correctional supervision.
Purpose of Drug Testing
At its core, drug testing is a method to identify drug use and track compliance with court-mandated conditions (Kilmer, 2008). Beyond the core surveillance function (Nguyen et al., 2021), there are two primary purposes for drug testing: (1) to deter substance use through threat of detection and sanctions (Paternoster, 2010) and (2) to inform treatment providers of progress in changing substance use behaviors. These two rationales for use of drug testing in a community supervision context will be discussed below.
Drug Testing as a Deterrent
The primary use of drug testing in community corrections is arguably as a deterrent (see Cullen et al., 2016). Deterrence theory assumes individuals have the capacity to make rational choices based on the costs and benefits of criminal engagement (Nagin, 2013). For drug testing to be effective as a deterrent, an agency must ensure a client receives a swift and certain sanction when drug use is detected, and graduated sanctions are applied for any subsequent positive tests (Hawken & Kleiman, 2009). As noted by Kilmer (2008), “the net benefit of using drugs will decrease when there is an increase in the expected cost associated with using them. This cost can increase when the probability of detection or the sanction for testing positive increases” (p. 95). Given that clients must be informed of the risk of drug testing for the deterrent effect to become active, it is possible that clients have incomplete information about drug testing (i.e., frequency, likelihood of being selected, etc.), thereby mitigating the potential deterrent effect due to a perceptual failure. Regardless, without drug testing, the ability of probation officers to detect drug use, thereby increasing the certainty of apprehension, would be extremely low.
Often referenced as “holding offenders accountable,” the use of drug tests to deter behavior assumes that sobriety can be achieved (at least in part) through accountability. According to Marlowe and Meyer (2011), accountability without punishment is ineffective and is therefore just a proxy for what is a deterrence model. Hence, a client using drugs violates their agreed-upon conditions and they must be held accountable for breaking the rules. A reliable drug testing system must be in place to test frequently and in close proximity to drug use so that a response can be applied swiftly and decisively (Torres, 1996). To meet this need, Marlowe and Meyer (2011) suggest that courts (via probation departments) adopt practices such as random, unpredictable, frequent drug testing, which is increased when drug use is detected and then followed by swift and certain punishment. The practices suggest that if the system can effectively monitor an individual’s drug use, and a punishment is administered, that compliance will be achieved. Given that even the best protocols will not detect all drug use among probationers, the question is whether a swift and certain response deters future drug use when drug use is detected.
The proposition that swift and certain punishment works was put to the test with the Honest Opportunity Probation with Enforcement (HOPE) program, initiated and tested in Hawaii, and replicated in sites across the United States. This model utilizes frequent randomized drug-testing to monitor and swiftly apply a graduated, non-severe sanction (e.g., a short jail stay) to drug use and other non-compliant behaviors (Lattimore et al., 2016). This model stands in contrast to traditional probation approaches that may require several failed drug tests before an individual sees a judge or is issued a violation. The HOPE program emphasizes a swift, certain, and fair (SCF) response to non-compliance, along with treatment alternatives for high-need individuals. An initial study using random assignment to either HOPE or probation as usual (PAU) found HOPE participants had fewer positive drug tests, fewer parole violations, and a decreased likelihood of arrest or incarceration for a new crime (Hawken & Kleiman, 2009). However, two subsequent randomized experiments failed to replicate these initial positive findings, showing instead similar recidivism outcomes between HOPE and PAU (Lattimore et al., 2016; O’Connell et al., 2016); this is consistent with findings from meta-analyses on deterrence-based correctional interventions that show limited treatment effects (Gendreau et al., 2000; Lipsey, 2009; McGuire, 2002).